基于AP加权的自适应动态室内定位方法

Adaptive dynamic indoor positioning method based on AP weighting

针对传统定位算法存在大量不同接入点(AP)的冗余信息,且在定位范围较大时定位时效性差的问题,提出一种基于卡方距离的AP加权自适应动态定位算法。指纹匹配阶段利用卡方距离代替传统加权K近邻算法(WKNN)中的欧式距离,用AP方差对相似度进行加权,从而自适应调整距离阈值,并引入动态K值,精确提取数据库信息。实验结果表明,与传统定位算法相比,该算法更有利于去掉冗余的AP,使定位误差范围缩小,可显著提高定位精度与稳定性。

In view of the redundant information of different Access Points(AP) in traditional location algorithm, and the time-effective error of location in the large location range, an AP weighted adaptive dynamic location algorithm based on the square distance is proposed. In fingerprint matching stage, the chi square distance is used to replace the European distance in the traditional Weighted K-Nearest Neighborhood(WKNN) algorithm, and the AP variance is weighted to the similarity, so as to adjust the distance threshold adaptively, and dynamic K value is introduced to extract the database information accurately. The experimental results show that the algorithm is more advantageous to remove redundant AP, reduce the range of positioning error and improve the positioning accuracy and stability.